on some Carnivorous Insects. 345 



drifted to the side every time I brought them back. I 

 did not see them tried. An E. hiarbas egg was picked up 

 and ejected, as was, several times, an egg of hippocoon 

 and one of P. demodocus; but a green egg, with dark 

 ring, of C. ethalion was very soon picked up and carried 

 mth the column. I then killed, diswanged and put just 

 inside the column a. ^ D. chrysippus and a r^ Acraea 

 natalica. Each was at once covered by a mass of ants and 

 gradually brought into the line. Proceedings continued 

 to be so slow that I had to leave, but on my return to 

 inspect half an hour later or less both had completely 

 disappeared. 



May 2nd. — Afternoon. 



This was really a larva experiment, but as eggs were 

 used too, and as the use of freshly-hatched larvae is also 

 obviously relevant, I ought perhaps to state the gist of 

 it here. 



Eggs of A. acara many times refused, one each of 

 hippocoon and P. demodocus two or three times ejected, 

 a just-hatched P. dardanus larva treated as in the morning, 

 but a three- quarter- grown one with the final, most protec- 

 tive appearance at once set upon and carried off. Barely 

 hatched A. acara larvae, still busy with their egg-shells, 

 persistently refused and repeatedly ejected, but a half- 

 grown individual of the same species killed and taken, as 

 also quarter-grown E. hiarbas and A. caldarena larvae and 

 a nearly full-grown A. caldarena larva. 



The larger Acraea larvae gave the ants a good deal of 

 trouble by exuding, when set upon, drops of the usual 

 poppy- flavoured liquid from the ends of their bristles. 

 They thus succeeded more than once in escaping from 

 the column. I put them back, however, and the ants 

 overcame the difficulty intentionally or incidentally by 

 placing on the ends of the bristles crumbs of dry earth, 

 which soaked up the liquid and enabled them to bite off 

 the bristles lower down. Fresh drops appeared as the 

 result of this, and fresh crumbs of earth were applied 

 until finally the bristles were razed off level, in many 

 cases, with the caterpillar's body. It was then set upon 

 freely by masses of ants, killed and carried off. I at first 

 beheved that the application of the earth-crumbs was 

 purely in the nature of " placarding " (which is still pos- 

 sible), but I felt before the end of the experiment that it 

 might readily be, as I have described it, for the purpose 



